In the unsettled waters of health care transformation, Dr. Kevin Volpp practices an alchemy of sorts: turning economic principles into incentives that lead to better health outcomes, individual wellness, and healthier living. If some money is saved in the process, we are all the better for it.
Dr. Volpp is a professor at the Perelman School of Medicine and the Wharton School at the University of Pennsylvania, where he is also Director of the Center for Health Incentives and Behavioral Economics.
His work has earned him the 2015 Matilda White Riley Award, issued by the National Institutes of Health Office of Behavioral and Social Sciences Research (OBSSR). The Matilda White Riley Award is given in recognition of an outstanding behavioral or social scientist whose research has contributed to both the deepening of knowledge and its application in a manner that furthers NIH’s mission of improving health.
When asked if he were to nominate himself for the award, what he would name as his major contribution, Dr. Volpp responds cheerfully: “That’s a difficult question!” Others, he thinks, would probably say he has led the translation of behavioral economics into health. However, he is the proudest about some very specific, measurable accomplishments.
“We have developed and tested financial incentives and smoking that have shown tripling of long-term smoking cessation rates among General Electric and CVS employees and that led to changes in employee benefit design in a number of large corporations,” said Dr. Volpp. “This has been important in helping to shift health insurance benefits from simply paying to treat disease to become more of a vehicle focused on keeping people healthy.”
Given the complexity and nature of some of these issues, the alchemy metaphor is rather apt. How should we think of behavioral economics?
Dr. Volpp would describe behavioral economics as a way of thinking about and viewing the world as it is applied both to understanding and influencing others’ behavior. “While people make a lot of decision errors, they tend to do so in systematic ways and, as I have written about at length with my colleague George Loewenstein, we can leverage many of these insights to develop interventions that help people.”
As Einstein would say of relativity, a theory that “was so beautiful it had to be true,” Dr. Volpp doesn’t recoil from framing the theoretical underpinnings of his work in terms of beauty.
“Behavioral economics describes many phenomena in terms of human behavior in ways that are both deep and elegant. I would highlight [Daniel] Kahneman and [Amos] Tversky’s prospect theory as, perhaps, the best example of this,” said Dr. Volpp, as he emphasizes the explicative points it comprises:
- The notion of reference dependence in which how a decision maker evaluates an outcome (as a gain or a loss) depends on his/her starting point;
- Loss aversion (the disutility of a loss is much stronger than the utility of an equivalent gain);
- Diminished sensitivity to both gains and losses depending on your starting point;
- Non-linear probability weighting (people overweigh small probabilities).
These theoretical pillars have helped Dr. Volpp think through the many practical challenges in the current environment of health care transformation and workplace wellness. What concerns him the most? What ideas on the role of incentives and behavioral economics are most misunderstood? In what areas are the hopes ahead of the science?
As part of the Affordable Care Act employers can provide rewards and penalties of up to 30-50 percent of total premiums, depending on outcome-based incentives such as: body mass index (BMI), blood pressure, cholesterol, and smoking. The total can be up to 50 percent if smoking is included, 30 percent if it is not, explains Dr. Volpp.
“There is a widely held assumption that incentives matter – which is true – but that the design doesn’t matter that much and that all that matters is the size of the reward – which is not true,” he said.
Many employers are implementing these programs by simply adjusting premiums for employees who do not meet threshold targets such as a BMI<25. This is unlikely to be effective for two main reasons, according to Dr. Volpp.
First, “people who have a BMI more than a little higher than 25 will be unlikely to feel they can reach the target and, [second], tying [the incentives] into premiums makes the rewards much less visible due to mental accounting,” he said.
In other words, people will readily notice an incremental shift from $0 to $500. However, if an employee makes $50,000 per year and the incentive would make his paycheck go up to $50,500, that same $500 will run a much higher risk of going unnoticed.
“Incentive programs designed in this manner could result in massive cost shifting to higher risk employees without much behavior change. It is critical to think about ways to make these types of incentive programs more effective at actually reducing risk/improving health by rewarding improvement or attainment of a threshold and by ‘unbundling’ , e.g. separating these rewards from premiums to make them more noticeable to employees and thereby more likely to change behavior,” emphasized Dr. Volpp. “The science of how to optimally do this is not yet where the policy makers have gone in making this part of standard benefit design for employers throughout the country.”
What areas of this research will live up to their promise soonest?
“As we have seen with the use of choice architecture and defaults transforming how retirement savings programs are run by employers, I think we will see a shift in how health systems modify physician order entry and how health plans present choices to patients, which will have similar effects,” said Dr. Volpp.
For example, rather than giving patients a long list of providers with variable cost and quality to choose from, “Why not set up a default to have the higher quality/lowest cost provider first or have providers order generics by default instead of expensive brand medications that are no more effective?,” the doctor asks. “These principles are also increasingly being used in food layout in cafeterias, menu layout and a variety of other areas to improve health.”
It is imperative to ask about the ethical limits to applying the principles of behavioral economics to health and health care. Dr. Volpp is categorical: “We have to be careful not to presume ‘we’ know what is best for people in all cases.”
“While no one is likely to object to putting the healthy food in the prime position in a cafeteria, picking the ‘appropriate’ default becomes more challenging in areas where, for example, there are cost/quality tradeoffs. Who decides what is appropriate? Who is ‘we’?
“There also are people who, with full knowledge of the present and future costs and benefits of their actions, may choose to continue to smoke or stay obese. As many others, such as Thaler and Sunstein, have described the ideal as ‘light paternalism’ where behavioral economic interventions are not overly heavy handed but rather make it easier for people to make healthy choices without requiring them to do so,” said Dr. Volpp.
The 8th Matilda White Riley Award and Lecture will take place June 23, 2015, as part of the celebrations of the 20th Anniversary of OBSSR. Asked, from his perspective and research, what pathways OBSSR should follow next, Dr. Volpp was not wanting for suggestions.
“The US ranks poorly on a variety of health metrics despite spending more than any country in the world. We need more research that addresses social and behavioral determinants of health and enables us to develop scalable interventions,” he said.
These interventions “can help change the paradigm of health care delivery from a reactive visit-based system to one that is much more proactive in managing population health. Advances in wireless technologies will help, but the technology alone will not solve these problems.”
This, he argues, will require shifting resources from innovation focused on developing new treatments into innovation focused on better managing disease, improving the efficiency of health care operations, improving health behavior, and influencing social determinants of health.
“We will need to map out how to do this efficiently in leveraging the capabilities and resources of providers, health plans and employers. We will need to figure out how to deliver much higher value in terms of population health to the American population as we can’t continue to spend far more than any other country but rank poorly on population health,” Dr. Volpp concluded.
PHOTO: Kevin Volpp (right) and colleagues David Asch (center), executive director of the Penn Medicine Center for Innovation, and Jingsan Zhu, assistant director of CHIBE.
PHOTO CREDIT: Penn Medicine, Summer 2013